2021 4th International Conference on Intelligent Autonomous Systems (ICoIAS) | 2021
Research on Recognition Method of Zinc Dross in Hot Dip Galvanizing Pot Based on Image Feature
Abstract
To realize the automatic removal of zinc slag in a hot-dip galvanizing pot, the recognition method of zinc slag based on features was studied. Three recognition algorithms of zinc slag based on image pixel value, SSIM (Structural Similarity Index), and image features are designed. MATLAB was used for experimental simulation, and the accuracy, precision, recall, F1- score, and execution efficiency of the three methods were compared. The results show that the zinc slag recognition method based on pixel value is more comprehensive, and the zinc slag recognition method based on statistical features has the highest accuracy, while the zinc slag recognition method based on SSIM has the best comprehensive effect in recognition.